Batch or continuous production: run length, changeover and the cost of stopping

Comparison criteria

Criteria are stated explicitly and neither option is declared a winner: which one fits depends on the constraint that binds hardest in your operation.

Changeover is a design parameter, not an overhead line

Plants running lots often treat setup as a fixed cost and respond by making longer runs, which inflates stock and lengthens the wait for anything not currently on the machine. The alternative response is to attack the setup itself: separating work that can be done while the asset still runs from work that genuinely requires it stopped, standardising fittings, staging materials and tools in advance, and removing adjustment by fixing positions. Every hour recovered widens the range of run lengths you can afford, which is what actually gives a lot-based plant its flexibility. Treating setup as immovable quietly converts a flexible asset into an inflexible one.

Uninterrupted running turns availability into the whole game

When a line is expected to run without stopping, the economics rest on how much of the available time it actually produces saleable output. An unplanned stop costs the outage, the material lost while conditions drift, and the time taken to bring quality back inside limits — often several times the duration of the fault itself. That reshapes priorities: maintenance becomes a scheduling problem rather than a repair function, spares for the constraining equipment are held whatever they cost to hold, and instrumentation is maintained as production equipment rather than as an accessory. Plants that budget only for the repair consistently underestimate what a stoppage removes.

Hybrid plants are normal and need a defined boundary

A great many operations run a continuous front end into a lot-based back end: bulk produced without interruption, then filled, packed or finished in discrete runs. The arrangement works well and creates one specific management problem, which is the interface. Bulk storage capacity between the two determines how independently they can run, and it is routinely undersized because it appears on a capital request as a tank rather than as flexibility. Decide deliberately how much buffer sits at that boundary, how bulk lots map onto packed lots for traceability, and who has authority to stop the front end when the back end is blocked.

Frequently asked questions

At what point does a lot-based plant justify moving to continuous running?

The signal is rarely a single volume threshold. Look instead at how much of your time goes into changeover and cleaning versus production, how narrow the product mix has become, and whether demand for the dominant product is steady enough to keep a dedicated asset loaded through a slow period. When one recipe already occupies most of the schedule and setup work consumes a large share of available hours, the case builds. Where the mix is still broadening, the same investment removes flexibility you are actively using.

How is traceability handled when material flows without natural lot breaks?

You define the boundaries yourself, usually by time window, by vessel, or by a marker introduced into the flow, and you record the process conditions and material inputs associated with each. The important part is deciding in advance how far a problem propagates, since material either side of a detected deviation may be affected. That decision governs how much you would have to hold or recover if a specification excursion is found later, so it deserves to be agreed with quality and commercial colleagues rather than set by convenience.

Can the same equipment be used for both modes?

Sometimes, and it is worth examining before committing capital. Equipment designed for uninterrupted service can often be run in campaigns, producing one grade for an extended period before switching, which recovers part of the flexibility while keeping the throughput. The limits are usually cleaning between grades, the volume of transitional material generated at each switch, and whether your control strategy can be re-tuned quickly. Campaign running is a genuine middle position rather than a compromise, provided the sequence is planned rather than improvised.